When is housing too expensive a hedge to hold?

I ended the last post saying “that the way I (and I think we) think of housing is better described as a hedge than as a punt. The price of the hedge is still important, but the goal for many is to be hedged so that they can stop worrying about their massive and unhedged liability.” What did I mean?

Well, if you don’t own a property and wouldn’t qualify for social housing, renting is a large, volatile and unavoidable expense. And owning your own home is a means to (almost) perfectly hedge the large, volatile and unavoidable expense of renting. If housing is a hedge and you own less housing than you will ultimately expect to be consuming (eg, you have a one-bed flat and are planning a large family) then you are basically short, rather than long, property. As the Bank said back in 2006 ‘when house prices rise, typically rents do too — so renters face higher housing costs’. So, if I live in an up and coming area, my rent will rise faster than average. We can’t stop and get off the UK housing market just because we’re not enjoying the ride.

But after the last post (about which Katie was characteristically kind) Tom Bowker posed a killer question:

Why is this a killer question? Two reasons.

  1. It is pertinent. As Tom suggests, it looks like I studiously avoided discussing any framework that might be used to work out whether it is a cheap or expensive hedge in my previous post on house prices.
  2. It is relevant beyond housing. It is a question that has been occupying almost every UK boardroom in the land. Not that they are obsessing on the price of housing as a hedge, but rather because they have been struggling on the price of hedging their defined benefit pension liabilities. And Tom’s question is about houses as hedges. As I suggested in the last blog, the London property market “make(s) me feel a bit like a de-risking defined benefit pension scheme”.

If we get are allowed to get a little geeky, the analogy of defined benefit pension schemes is not as tenuous as it might sound:

  • Pension scheme sponsors are short a stream of long-dated future debt-like cash-flows that will vary in line with the product of inflation, demographic factors and an idiosyncratic factor related to their scheme membership. There is a structural shortage of hedging instruments (long-dated inflation-linked bonds) given lack of supply and foreign demand.
  • Renters are short a stream of long-dated future debt-like cash-flows that will vary in line with the product of aggregate rental inflation, demographic factors, and an idiosyncratic factor related to the property in which they live. There is a structural shortage of hedging instruments (houses) given lack of supply and foreign demand.

As such, both renters and pension scheme sponsors are at risk of fluctuations in some combination of long-term real interest rates, a form of inflation, some demographic considerations, and some idiosyncratic factors. And their ability to hedge appears frustrated by structural supply-demand imbalances. Furthermore, it turns out that the way in which UK boardrooms monitor the price of their pension hedge can be used to monitor the price of housing as a hedge.

Long-dated inflation-linked, conventional Gilt yields plus a mix of high-yield and investment grade credit spreads vs UK residential rental yields*

housing ditty

  Pension schemes has had their regulatory incentives changed meaningfully over the past couple of decades and this has prompted a good deal of hedging. How can we monitor the changing price of the hedge to defined benefit pension scheme sponsors? I would suggest a combination of long-dated inflation-linked yields, long-dated conventional gilt yields and some combination of investment grade and high yield credit spreads (to account for a rough measure of inflation risk, long-dated real rate risk, and market risk premia that might be employed by schemes to match liabilities). And how can we monitor the changing price of the hedge to renters? I would suggest the residential rental yield. Both are shown in the chart above. It don’t think that it is unreasonable to suggest that attempts by pensions to hedge their liabilities have impacted the price of the hedge available to renters.

I believe that hedging activity in the pensions world has changed the economics of hedging in the housing world. Because pension funds have increasingly wanted to hedge their (plain vanilla) inflation risk (but there has been insufficient supply of hedging instruments), the cost of hedging inflation has become very expensive. This manifests in suppressed long-dated real yields (boosting the price of long-dated inflation-linked bond prices), and has contributed to a rise in price of other forms of inflation hedging (like housing). I should be clear at this point that connecting pension hedging to house prices is (today) a minority position. What is the upshot of this minority view? Well, although housing looks (to me) overvalued as a punt given my expectation of a rise in debt-service ratios over the next couple of years (eg it is a hedge that fewer people will be able to afford if rates rise, dampening demand), it doesn’t look like the most expensive hedge out there given changes in cost of other forms of inflation hedging (for those with the means to hedge). And there’s the tax incentive too (eg, as an owner you pay yourself rent out of your gross income, as a renter you pay someone else rent out of your net income). Which is not insubstantial.



* The residential rental yield is pulled together by taking the product of the average Nationwide house price in 2012 and Countrywide’s rental survey yield for 2012 as an anchor, and then the projecting this average rental forwards and backwards in time using the rental subcomponent of the RPI. This was then divided by the Nationwide house price time series to create a rental yield. It is massively imperfect and I’d be delighted if someone wanted to send me a better time series. The rental yield looks quite high to me, but this (I suspect) is because I live in London. In London the tax incentive to buy will be stronger given higher marginal tax rates associated with higher incomes. And it is also the market in which some folk look to hedge their domestic political risk (eg wealthy foreigners living in unstable regimes seeking a bolt-hole where the rule of law is strong).


The UK housing market is apparently not out of hand

Given that we are still arguing over the causes of the Great Depression, the First World War, and the French Revolution it is perhaps too much to ask that we should have a clear understanding as to the causes of the Global Financial Crisis.

But the popular narrative rightly had housing in the mix. Grand-scale liquidity mis-match in banks and shadow banks, megalomania on the part of financial leaders, large-scale leverage, and copious use of synthetic credit – sure. But housing was definitely somewhere in there.

The Bank of England understands the centrality of housing to the UK financial system. If it didn’t it wouldn’t have its Governor talk about housing as the biggest risk to financial stability, or put out wonderful charts like this one:

balance sheet fsr jun-14

But in releasing new macroprudential rules to govern mortgage lending in order to stop things getting out of hand, the Bank has basically said not only that things are not out of hand today, but also makes it the official position of the Bank that the market never got out of hand during 2006-2007. I have annotated in red a couple more charts from the Bank’s chartastic Financial Stability Report to show what I mean: a ceiling of 15% of mortgage lending to mortgages that are 4.5X income is out there in uncharted territory.

Out f hand FSR jun-14

As a Londoner, I struggle to ex-post rationalise why thing are not already out of hand (although remind myself that London is not the UK). In Haringey where I live the median property is priced at around £355k, which is 10.5X median earnings for the borough. And while that’s towards the upper end of the scale of area-specific earning-multiples, it’s not at the top of it. Inner London more generally is priced at a multiple of just over 10 times Inner London earnings (which are much higher than the national average).

Ratio of lower quartile/median house price to lower quartile/median earnings, England and Inner London, 1997-2013


Source: DLGC, data here

Sure, it is not the Bank’s job to fix house prices. But wow! Price-to-incomes look white hot in a way that hasn’t typically ended well.

Of course, price-to-income ratios are just one valuation measure of housing. Another popular one is the debt service ratio (eg, the proportion of disposable income spent on debt service of which mortgage payments make up the lion’s share). The following chart and scenario projections come from Credit Suisse.


On this measure housing has rarely looked cheaper. But with the economy motoring, unemployment falling fast, and inflation ticking up, it is not hard to see why the forward path of interest rates is signalling a number of interest rate increases over the next 18 months that will increase the cost of servicing a mortgage and removing a crutch that many observers believe is supporting the market.

So, that causes house prices to fall, right? Maybe. But here’s the thing.

I am fortunate enough to have been able to buy a property a few years ago. And I’m not selling my property even though I think that the bit of the market in which I live is out of hand. The biggest reason is non-financial (I value my marriage), but there are financial reasons too. These financial reasons make me feel a bit like a de-risking defined benefit pension scheme. That is to say that the way I (and I think we) think of housing is better described as a hedge than as a punt. The price of the hedge is still important, but the goal for many is to be hedged so that they can stop worrying about their massive and unhedged liability. This may make immediate sense to some, but I will endeavour to explain myself in my next post.